Limits...
Rapid Computational Prediction of Thermostabilizing Mutations for G Protein-Coupled Receptors.

Bhattacharya S, Lee S, Grisshammer R, Tate CG, Vaidehi N - J Chem Theory Comput (2014)

Bottom Line: Deriving thermostable mutants has been a successful strategy to stabilize GPCRs in detergents, but this process is experimentally tedious.The prediction using the stability score improves when using an ensemble of receptor conformations compared to a single structure, showing that receptor flexibility is important.We improved the thermostability prediction by including other properties such as residue-based stress and the extent of allosteric communication by each residue in the stability score.

View Article: PubMed Central - PubMed

Affiliation: Division of Immunology, Beckman Research Institute of the City of Hope , 1500 East Duarte Rd, Duarte, California 91010, United States.

ABSTRACT
G protein-coupled receptors (GPCRs) are highly dynamic and often denature when extracted in detergents. Deriving thermostable mutants has been a successful strategy to stabilize GPCRs in detergents, but this process is experimentally tedious. We have developed a computational method to predict the position of the thermostabilizing mutations for a given GPCR sequence. We have validated the method against experimentally measured thermostability data for single mutants of the β1-adrenergic receptor (β1AR), adenosine A2A receptor (A2AR) and neurotensin receptor 1 (NTSR1). To make these predictions we started from homology models of these receptors of varying accuracies and generated an ensemble of conformations by sampling the rigid body degrees of freedom of transmembrane helices. Then, an all-atom force field function was used to calculate the enthalpy gain, known as the "stability score" upon mutation of every residue, in these receptor structures, to alanine. For all three receptors, β1AR, A2AR, and NTSR1, we observed that mutations of hydrophobic residues in the transmembrane domain to alanine that have high stability scores correlate with high experimental thermostability. The prediction using the stability score improves when using an ensemble of receptor conformations compared to a single structure, showing that receptor flexibility is important. We also find that our previously developed LITiCon method for generating conformation ensembles is similar in performance to predictions using ensembles obtained from microseconds of molecular dynamics simulations (which is computationally hundred times slower than LITiCon). We improved the thermostability prediction by including other properties such as residue-based stress and the extent of allosteric communication by each residue in the stability score. Our method is the first step toward a computational method for rapid prediction of thermostable mutants of GPCRs.

No MeSH data available.


Related in: MedlinePlus

(A, B; E, F) Thermostableresidues that were identified using differentGPCR structures: (A) β1AR; (B) A2AR; (E)active-like state crystal structure of NTSR1; (F) homology model ofinactive NTSR1 based on β2AR. Experimental thermostability scoreshave been normalized so that the experimental value for the wild typereceptor is 50%. If a residue was correctly identified by a particularmodel, the cell corresponding to that model is colored orange; theresidues that were identified by all models are colored green, onesthat were identified only by the crystal structures or close templatehomology models are colored yellow; those that were not identifiedby any of the models are colored red. (C, D; G, H) crystal structuresof (C) β1AR, (D) A2AR, (G) NTSR1 +NT,and (H) NTSR1 −NT showing the thermostable mutation positions.+NT and −NT refer to NTSR1 mutants that are thermostable inpresence and absence of neurotensin, respectively.
© Copyright Policy
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC4230369&req=5

fig8: (A, B; E, F) Thermostableresidues that were identified using differentGPCR structures: (A) β1AR; (B) A2AR; (E)active-like state crystal structure of NTSR1; (F) homology model ofinactive NTSR1 based on β2AR. Experimental thermostability scoreshave been normalized so that the experimental value for the wild typereceptor is 50%. If a residue was correctly identified by a particularmodel, the cell corresponding to that model is colored orange; theresidues that were identified by all models are colored green, onesthat were identified only by the crystal structures or close templatehomology models are colored yellow; those that were not identifiedby any of the models are colored red. (C, D; G, H) crystal structuresof (C) β1AR, (D) A2AR, (G) NTSR1 +NT,and (H) NTSR1 −NT showing the thermostable mutation positions.+NT and −NT refer to NTSR1 mutants that are thermostable inpresence and absence of neurotensin, respectively.

Mentions: We have tested the performance and sensitivity of the LITiConDesignmethod by using homology models of varying resolution (closeness tothe crystal structure). We find that the homology models derived usingclose homologues as templates (templates with low RMSD to the targetcrystal structure) perform better than the ones based on distant templates.Parts a and bo of Figure 8 show the performanceof the different homology models of β1AR and A2AR, respectively. Parts e and f of Figure 8 show the performance of the NTSR1 models. For the +NT data,we used the crystal structure of active NTSR1, while for the −NTdata, we used the homology model of inactive NSTR1 based on the β2AR crystal structure. We calculated the number of true thermostablemutants that are recovered in the top 50% when sorted by calculatedthermostability scores. In Figure 8a and b,for each model, the residues that are correctly identified are highlightedin orange. As expected, for β1AR, the crystal structure,and the homology model of β1AR based on β2AR and D3DR performed better than the distant A2AR and CXCR4 based models. Residues that are correctly identifiedby most of the structures are colored green, the ones that are identifiedby only the close template models and crystal structures are coloredyellow, while the ones that are not identified by any of the modelsare colored red. For the NTSR1 models (Figure 8e and f), residues that are correctly identified by the crystal structureor homology model are colored green, and the ones that are not identifiedare colored red. These residues are also highlighted in Figure 8c, d, g, and h. For β1AR, the threemutants (I551.46A, V902.47A, I1293.40A) that are not identified by any of the models are all located intightly packed regions of the receptor, facing the core of the TMdomain. In contrast, the residues that are successfully predictedby our method are located in loosely packed regions of the receptor(near the extracellular or intracellular termini of the TM helices)or facing the lipid bilayer. For the residues located in tightly packedregions of the receptor, the side chain optimization methods are inadequateto account for the repacking of the side chains upon mutation. Thiscould be a possible reason for the failure of our method in not identifyingcertain residues. Also, the lack of electrostatic component in ourenergy function could contribute to the failure of this method inidentifying certain mutations. The electrostatic energy could be importantfor mutations in tightly packed regions of the receptor that are inthe neighborhood of polar residues. It is difficult to accuratelyestimate the electrostatic energy, especially in a hybrid environmentsuch as the lipid bilayer. We have excluded the electrostatic componentfrom our scoring function, since adding the electrostatic energy worsenedthe predictability of the thermostabilizing mutations. However, specializedenergy functions for hydrogen bonds/salt bridges could be developedto model the effect of the polar residues in modulating receptor stability.Such energy functions are currently under development in our lab.


Rapid Computational Prediction of Thermostabilizing Mutations for G Protein-Coupled Receptors.

Bhattacharya S, Lee S, Grisshammer R, Tate CG, Vaidehi N - J Chem Theory Comput (2014)

(A, B; E, F) Thermostableresidues that were identified using differentGPCR structures: (A) β1AR; (B) A2AR; (E)active-like state crystal structure of NTSR1; (F) homology model ofinactive NTSR1 based on β2AR. Experimental thermostability scoreshave been normalized so that the experimental value for the wild typereceptor is 50%. If a residue was correctly identified by a particularmodel, the cell corresponding to that model is colored orange; theresidues that were identified by all models are colored green, onesthat were identified only by the crystal structures or close templatehomology models are colored yellow; those that were not identifiedby any of the models are colored red. (C, D; G, H) crystal structuresof (C) β1AR, (D) A2AR, (G) NTSR1 +NT,and (H) NTSR1 −NT showing the thermostable mutation positions.+NT and −NT refer to NTSR1 mutants that are thermostable inpresence and absence of neurotensin, respectively.
© Copyright Policy
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC4230369&req=5

fig8: (A, B; E, F) Thermostableresidues that were identified using differentGPCR structures: (A) β1AR; (B) A2AR; (E)active-like state crystal structure of NTSR1; (F) homology model ofinactive NTSR1 based on β2AR. Experimental thermostability scoreshave been normalized so that the experimental value for the wild typereceptor is 50%. If a residue was correctly identified by a particularmodel, the cell corresponding to that model is colored orange; theresidues that were identified by all models are colored green, onesthat were identified only by the crystal structures or close templatehomology models are colored yellow; those that were not identifiedby any of the models are colored red. (C, D; G, H) crystal structuresof (C) β1AR, (D) A2AR, (G) NTSR1 +NT,and (H) NTSR1 −NT showing the thermostable mutation positions.+NT and −NT refer to NTSR1 mutants that are thermostable inpresence and absence of neurotensin, respectively.
Mentions: We have tested the performance and sensitivity of the LITiConDesignmethod by using homology models of varying resolution (closeness tothe crystal structure). We find that the homology models derived usingclose homologues as templates (templates with low RMSD to the targetcrystal structure) perform better than the ones based on distant templates.Parts a and bo of Figure 8 show the performanceof the different homology models of β1AR and A2AR, respectively. Parts e and f of Figure 8 show the performance of the NTSR1 models. For the +NT data,we used the crystal structure of active NTSR1, while for the −NTdata, we used the homology model of inactive NSTR1 based on the β2AR crystal structure. We calculated the number of true thermostablemutants that are recovered in the top 50% when sorted by calculatedthermostability scores. In Figure 8a and b,for each model, the residues that are correctly identified are highlightedin orange. As expected, for β1AR, the crystal structure,and the homology model of β1AR based on β2AR and D3DR performed better than the distant A2AR and CXCR4 based models. Residues that are correctly identifiedby most of the structures are colored green, the ones that are identifiedby only the close template models and crystal structures are coloredyellow, while the ones that are not identified by any of the modelsare colored red. For the NTSR1 models (Figure 8e and f), residues that are correctly identified by the crystal structureor homology model are colored green, and the ones that are not identifiedare colored red. These residues are also highlighted in Figure 8c, d, g, and h. For β1AR, the threemutants (I551.46A, V902.47A, I1293.40A) that are not identified by any of the models are all located intightly packed regions of the receptor, facing the core of the TMdomain. In contrast, the residues that are successfully predictedby our method are located in loosely packed regions of the receptor(near the extracellular or intracellular termini of the TM helices)or facing the lipid bilayer. For the residues located in tightly packedregions of the receptor, the side chain optimization methods are inadequateto account for the repacking of the side chains upon mutation. Thiscould be a possible reason for the failure of our method in not identifyingcertain residues. Also, the lack of electrostatic component in ourenergy function could contribute to the failure of this method inidentifying certain mutations. The electrostatic energy could be importantfor mutations in tightly packed regions of the receptor that are inthe neighborhood of polar residues. It is difficult to accuratelyestimate the electrostatic energy, especially in a hybrid environmentsuch as the lipid bilayer. We have excluded the electrostatic componentfrom our scoring function, since adding the electrostatic energy worsenedthe predictability of the thermostabilizing mutations. However, specializedenergy functions for hydrogen bonds/salt bridges could be developedto model the effect of the polar residues in modulating receptor stability.Such energy functions are currently under development in our lab.

Bottom Line: Deriving thermostable mutants has been a successful strategy to stabilize GPCRs in detergents, but this process is experimentally tedious.The prediction using the stability score improves when using an ensemble of receptor conformations compared to a single structure, showing that receptor flexibility is important.We improved the thermostability prediction by including other properties such as residue-based stress and the extent of allosteric communication by each residue in the stability score.

View Article: PubMed Central - PubMed

Affiliation: Division of Immunology, Beckman Research Institute of the City of Hope , 1500 East Duarte Rd, Duarte, California 91010, United States.

ABSTRACT
G protein-coupled receptors (GPCRs) are highly dynamic and often denature when extracted in detergents. Deriving thermostable mutants has been a successful strategy to stabilize GPCRs in detergents, but this process is experimentally tedious. We have developed a computational method to predict the position of the thermostabilizing mutations for a given GPCR sequence. We have validated the method against experimentally measured thermostability data for single mutants of the β1-adrenergic receptor (β1AR), adenosine A2A receptor (A2AR) and neurotensin receptor 1 (NTSR1). To make these predictions we started from homology models of these receptors of varying accuracies and generated an ensemble of conformations by sampling the rigid body degrees of freedom of transmembrane helices. Then, an all-atom force field function was used to calculate the enthalpy gain, known as the "stability score" upon mutation of every residue, in these receptor structures, to alanine. For all three receptors, β1AR, A2AR, and NTSR1, we observed that mutations of hydrophobic residues in the transmembrane domain to alanine that have high stability scores correlate with high experimental thermostability. The prediction using the stability score improves when using an ensemble of receptor conformations compared to a single structure, showing that receptor flexibility is important. We also find that our previously developed LITiCon method for generating conformation ensembles is similar in performance to predictions using ensembles obtained from microseconds of molecular dynamics simulations (which is computationally hundred times slower than LITiCon). We improved the thermostability prediction by including other properties such as residue-based stress and the extent of allosteric communication by each residue in the stability score. Our method is the first step toward a computational method for rapid prediction of thermostable mutants of GPCRs.

No MeSH data available.


Related in: MedlinePlus